Skip to content

llmproxy documentation

A self-hosted, OpenAI-compatible LLM proxy: one static Go binary, per-user API keys, a curated model catalog and usage accounting. No request or response content is ever persisted.

Start here

I want to Go to
See it work, end to end Getting started
Deploy it on a server Install the binary
Deploy it with containers Install with Docker
Give my team access Set up SSO
Meter Claude Code Meter Claude Code
Look up a variable or endpoint Configuration, API
Understand why it is built this way Architecture

Deploy it

Operate it

Look it up

Understand it

  • concepts/architecture.md: scope and non-goals, the endpoint families, and every design rule with its rationale, including the no-content-persistence guarantee.

Contribute

  • CONTRIBUTING.md: development setup, the writing rules for these pages, the commit convention, how a release is cut.
  • SECURITY.md: how to report a vulnerability privately.